Pluralistic and stochastic gene regulation: examples, models and consistent theory.

Area of Science:

• Molecular Biology
• Systems Biology
• Genomics

Background:

• Gene regulation is often modeled simplistically, not accounting for observed biological variability.
• Evidence suggests transcription factor (TF) levels, binding, and activity fluctuate, and gene expression occurs in bursts.
• Evolutionary analysis shows frequent loss/gain of regulators and rewiring of regulatory sites, indicating dynamic processes.

Purpose of the Study:

• To develop a theory of pluralistic and stochastic gene regulation.
• To integrate empirical data with mathematical modeling for a comprehensive understanding.
• To investigate the regulatory mechanisms of co-expressed gene pairs, including ribosomal proteins.

Main Methods:

• Meta-analysis of ENCODE ChIP-Seq experiments to study TF binding.
• Analysis of approximately 800,000 co-expressed human gene pairs.
• Machine learning models to predict transcript levels based on TF binding sites.

Main Results:

• Widespread pluralistic regulation observed in ≈800,000 co-expressed gene pairs.
• Approximately 50% of observed regulators bind reproducibly to both genes in a pair, twice as often as in independently expressed pairs.
• No single master regulator identified; machine learning accurately predicted transcript levels despite TF binding fluctuations.

Conclusions:

• Gene regulation is inherently pluralistic and stochastic, characterized by partially random binding, redundancy, and burst-like expression.
• The theory explains degeneracy of binding motifs and massive regulatory rewiring during evolution.
• Findings challenge traditional views and provide a more realistic framework for gene regulation research.